AIDS-THE GREAT NUMBERS GAME
By Peter Plumley
Rethinking AIDS Oct./Nov. 1993
The AIDS epidemic has presented a unique opportunity for special interest groups
to manipulate numbers. Furthermore, although there are few people interested in
downplaying the epidemic, there are many who would like to exaggerate its impact.
Some want to use it as an excuse to discourage non-marital sex. Medical
researchers want more AIDS research funding. Public health officials don't want to
risk being accused of complacency about a major infectious disease that has quite
properly attracted worldwide attention. Gay activists want people to feel that it is
everyone's disease-not just a gay problem, so that more funding will be available.
Politicians realize that they can't lose by engaging in the "politically correct"
business of promoting "AIDS awareness" and "AIDS prevention."
Finally, the media understands that the public will read any story about AIDS,
particularly if it has numbers that get attention. Because few in the media really
understand the numbers, the special interest groups are able to manipulate the data
to put the epidemic in its worst light. Here are some of the tricks used in "the great
"We are all at risk!" True enough. But the risk varies-from very high for the IV
drug user who shares needles with his sick friends, to too remote to be of concern
for most of us. In fact, for most people, the risk is so low that worrying about it
may cause more harm than good because of the extra stress that results.
"The sky is falling!" Nearly all of the estimates of HIV+ people and projections of
future AIDS cases made by public health officials have turned out to be too high.
This is partly because of a lack of understanding of the nature of the epidemic; but
it also occurs because (1) no one wants to appear complacent; (2) the worse the
epidemic is projected to be, the more money will be made available for public
health work; and (3) when the numbers turn out to be lower, the officials can take
credit for having done a good job of AIDS education.
Use of cumulative numbers. A headline reported that one group had a "70%
increase in AIDS cases in the past two years." Actually, the number of new cases
was less than two years earlier-but the cumulative number of cases since the
epidemic started was up 70%. (Using this logic, every disease is on the increase.)
Unfortunately, this technique of presenting AIDS data is all too common, and
frequently misleads even the better reporters.
Use of percentage increases for small groups. If the number of cases of a disease
goes from 1000 to 1500 for one group of people, and from 10 to 20 for another
group, as a public health matter it would seem that the first group needed the most
attention. Yet frequently we hear that the second group "had the highest rate of
increase in cases," because it had a 100% increase, while the first group had only a
"Russian Roulette" and other fallacies. How often have we heard that "engaging in
unprotected sex is like playing Russian Roulette?" Yet with Russian Roulette there
is a one chance in six of taking the fatal bullet, whereas for most people, there is
only a remote chance of sexual contact with an HIV-positive person, and only a
very small chance of infection even if one's partner is infected, because HIV is so
extremely difficult to transmit by penile-vaginal sex. Thus those who use the
"Russian Roulette" warning are equating a one-in-six risk with a risk of
"one-in-a-million" or less.
"When you have sex with someone, you're having sex with all of his partners, and
his partner's partners, and their partners, etc." The implication is that if any of these
previous generations of sexual partners was HIV-positive, you'll become infected.
The statement implies a near-100% efficiency of transmission of HIV by
penile-vaginal sex. Many articles and warnings have implicitly assumed that sexual
contact with an HIV+ person automatically means that the partner will become
infected, and polls indicate that most people believe that sexual contact with an
infected person is likely to result in HIV infection. In fact, it is so difficult to
transmit HIV in this manner that one's prior heterosexual partners rarely ever
matter for heterosexuals. (However, they do make a difference for homosexuals,
because of the greater efficiency of transmission of HIV by anal sex.)
Failure to put numbers in perspective. Another headline said "AIDS has become
the fifth leading cause of death for people ages 15-40." Sounds serious-until one
realizes that (1) overall death rates are very low at the younger ages, and (2) most
of the deaths that do occur are concentrated in the top four causes.
Overemphasis on isolated, "freak" occurrences of HIV transmission. The
transmission of HIV by a Florida dentist to five patients was an aberration, caused
by either a major failure to observe normal dental procedures, or a deliberate act of
murder. Yet the media treated it as if it represented a new-and possibly
common-means of transmission, and millions of people began to fear a trip to the
"Save the Children!" We all want to protect our young people from harm. The
AIDS epidemic has given the special interest groups the chance to use this desire to
push for greater prevention efforts. We constantly hear about the "exploding AIDS
epidemic" among teenagers. Yet the number of AIDS cases reported to the CDC
for ages 13-19 declined from 170 in 1990 to 159 in 1992. For ages 20-24
(representing people who probably became infected with HIV while teenagers), the
decline was from 1626 cases in 1990 to 1446 in 1992.
For ages 13-24, AIDS cases attributed to heterosexual contact have increased
somewhat, from 243 in 1990 to 295 in 1992. This was 7% of the total of 4,114
AIDS cases reported as arising from heterosexual contact in 1992 for all ages
"I was protected-I was wearing a condom!" There was the drug addict who was
warned not to share needles, because he might get AIDS. He responded that it was
okay because he was wearing a condom. Yet, how many times have we heard total
numbers of AIDS cases quoted, particularly for teenagers, and then used as a
reason for AIDS education campaigns aimed at prevention of heterosexual
And finally, there are simple mistakes in numbers. These can happen in any article,
but happen more often with AIDS (and are more apt to be believed by the reader)
because it is such a complicated disease. This author has discovered many such
errors, including in such distinguished publications as the Journal of the American
Medical Association, where an article overstated a rate of HIV infection by a
factor of 100 because of confusion over the mathematics of percentages, and the
headline of a table in the 1991 issue of the World Almanac and Book of Facts
which claims that the table lists "U.S. Metropolitan Areas with AIDS Rates of 25%
or more." The actual list is for cities with rates of 25 per 100,000 or more, not 25
Peter W. Plumley is the author of "Modeling the AIDS Epidemic by Analysis of Sexual and
Intravenous Drug Behavior" and "An Analysis of the AIDS Epidemic as it Affects
Heterosexuals," both published in the Transactions of the Society of Actuaries.